Agents

An agent file should be more than a boundary list.

Signalane treats agent instructions as collaboration architecture: anchor, role, current truth, evidence, handoff, and recovery path.

Mini thesis

Agents do not stay reliable because a file says "do this" and "do not do that."

Boundary lists stop obvious mistakes, but serious agent work fails through quieter drift: stale scope, lost reasoning, preserved handoffs that no longer match the work, and instructions that cannot return to a living decision point.

Read first The Anchor Is Not the File

Stable AI collaboration depends on a living return point, not only instruction files.

Deepen Anchored Agent Files and Collaboration Constitution

A public-safe method note on anchored, person-specific operating files.

Operational safety Handoffs Are Safety Surfaces

Continuity needs truth, not performance, when work crosses sessions or agents.

Next branch: Agentic AI